← Back to community benchmarks
Qwen3.5-4B
Performance
32k
tokens
576.0
PP tok/s
28.6
TG tok/s
56893
TTFT (ms)
6.9
Peak mem (GB)
Hardware
Chip
M2 Max (38c)
Memory
64 GB
GPU Cores
38
Software
oMLX
v0.2.6
macOS
macOS 15.5
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 736.4 | 102.2 | 4.0 GB | view |
| 4k | 734.5 | 95.4 | 4.2 GB | view |
| 8k | 720.3 | 87.0 | 4.6 GB | view |
| 16k | 691.3 | 73.4 | 5.3 GB | view |
| 32k | 576.0 | 28.6 | 6.9 GB | current |
| 64k | 468.7 | 18.2 | 8.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 102.2 | 1.00× |
| 2× | 80.1 | 0.78× |
| 4× | 91.2 | 0.89× |
| 8× | 78.4 | 0.77× |